HYPERDIMENSIONAL COMPUTING DEVICE
    1.
    发明公开

    公开(公告)号:US20240274199A1

    公开(公告)日:2024-08-15

    申请号:US18166484

    申请日:2023-02-09

    CPC classification number: G11C16/08 G11C16/0483 G11C16/26

    Abstract: A hyperdimensional computing device includes a non-volatile memory cell array and a first operation circuit. The non-volatile memory cell array is coupled to a plurality of first word lines. The non-volatile memory cell array has a plurality of memory cell groups, a plurality of first memory cells of each of the memory cell groups are coupled to a same first word line of the first word lines, and the memory cell groups respectively store a plurality of data vectors. The first operation circuit receives at least one of the data vectors through bit lines and generates a bundled data vector according to the at least one of the data vectors.

    MULTIPLICATION AND ADDITION OPERATION DEVICE AND CONTROL METHOD FOR MULTIPLICATION AND ADDITION OPERATION THEREOF

    公开(公告)号:US20220236951A1

    公开(公告)日:2022-07-28

    申请号:US17344500

    申请日:2021-06-10

    Abstract: A multiplication and addition operation device and a control method thereof are provided. The multiplication and addition operation device includes a feature information filter and an in-memory calculator. The feature information filter records a plurality of designated bits of a plurality of feature information, compares received input information with the designated bits to generate a comparison result, and generates a selected address according to the comparison result. The in-memory calculator records all bits of the feature information, and generates an operation result by performing a multiplication and addition operation on the feature information and the input information according to the selected address.

    Semiconductor circuit and operating method for the same

    公开(公告)号:US10971200B2

    公开(公告)日:2021-04-06

    申请号:US16154831

    申请日:2018-10-09

    Abstract: A semiconductor circuit and an operating method for the same are provided. The method includes the following steps. A memory circuit is operated during a first timing to obtain a first memory state signal S1. The memory circuit is operated during a second timing after the first timing to obtain a second memory state signal S2. A difference between the first memory state signal S1 and the second memory state signal S2 is calculated to obtain a state difference signal SD. A calculating is performed to obtain an un-compensated output data signal OD relative with an input data signal ID and the second memory state signal S2. The state difference signal SD and the un-compensated output data signal OD are calculated to obtain a compensated output data signal OD′.

    Multi-state memory device and method for adjusting memory state characteristics of the same

    公开(公告)号:US10482953B1

    公开(公告)日:2019-11-19

    申请号:US16103022

    申请日:2018-08-14

    Abstract: A multi-state memory device includes a first memory element, a second memory element, a first controlling element and a second controlling element. The second memory element has a memory cell structure identical to that of the first memory element and connects to the first memory element in series. The first controlling element is connected to the first memory element either in series or in parallel. The second controlling element has a characteristic value identical to that of the first controlling element and is connected to the second memory element by a connection structure identical to that of the first controlling element. When the first memory element receives a first signal and a second signal through the first controlling element, a first state value and a second state value are generated correspondingly, and the characteristic value is greater than the first state value and less than the second state value.

    Memory and training method for neural network based on memory

    公开(公告)号:US12182701B2

    公开(公告)日:2024-12-31

    申请号:US17388053

    申请日:2021-07-29

    Abstract: The present invention discloses a memory and a training method for neural network based on memory. The training method includes: obtaining one or more transfer functions of a memory corresponding to one or more influence factors; determining a training plan according to an ideal case and the one or more influence factors; training the neural network according to the training plan and the one or more transfer functions to obtain a plurality of weights of the trained neural network; and programming the memory according to the weights.

Patent Agency Ranking